Cardiac Imaging to Assess Left Ventricular Systolic Function in Atrial Fibrillation

The validity and reproducibility of systolic function assessment in patients with atrial fibrillation (AF) using cardiac magnetic resonance, echocardiography, nuclear imaging and computed tomography is unknown. A prospectively-registered systematic review was performed, including 24 published studies with patients in AF at the time of imaging and reporting validity or reproducibility data on left ventricular systolic parameters (PROSPERO: CRD42018091674). Data extraction and risk of bias were performed by 2 investigators independently and synthesized qualitatively. In 3 cardiac magnetic resonance studies (40 AF patients), left ventricular ejection fraction and stroke volume measurements correlated highly with catheter angiography (r ≥0.85), and intra- and/or interobserver variability were low. From 3 nuclear studies (171 AF patients), there were no external validation assessments but intra and/or interobserver and intersession variability were low. In 18 echocardiography studies (2,566 AF patients), 2 studies showed high external validity of global longitudinal strain and tissue Doppler s' with angiography-derived dP/dt (r ≥0.88). Global longitudinal strain and myocardial performance index were both associated with adverse cardiovascular events. Reproducibility of echocardiography was better when selecting an index-beat (where 2 preceding R-to-R intervals are similar) compared to averaging of consecutive beats. There were no studies relating to computed tomography. Most studies were small and biased by selection of patients with good quality images, limiting clinical extrapolation of results. The validity of systolic function measurements in patients with AF remains unclear due to the paucity of good-quality data.